System and method for automatically recognizing delivery point information

Patent metadata

Field	Value
Publication number	US-12387476-B2
Application number	US-202418627368-A
Country	US
Kind code	B2
Filing date	Apr 4, 2024
Priority date	Apr 1, 2020
Publication date	Aug 12, 2025
Grant date	Aug 12, 2025

Abstract

This application relates to a system for automatically recognizing geographical area information provided on an item. The system may include an optical scanner configured to capture geographical area information provided on an item, the geographical area information comprising a plurality of geographical area components. The system may also include a controller in data communication with the optical scanner and configured to recognize the captured geographical area information by running a plurality of machine learning or deep learning models separately and sequentially on the plurality of geographical area components of the captured geographical area information.

First claim

What is claimed is:

1. A system for sorting items, the system comprising:
an optical scanner configured to capture destination information on an item, the destination information comprising a first geographical component and a second geographical component;
a memory configured to store a plurality of machine learning models, wherein the plurality of machine learning models comprises:
a first machine learning model trained to recognize the first geographical component from the captured destination information, and
a second machine learning model trained to recognize the second geographical component from the captured destination information based at least in part on the recognized first geographical component; and

one or more processors in data communication with the optical scanner, the one or more processors configured to:
recognize the first geographical component from the captured destination information using the first machine learning model;
based on recognizing the first geographical component, identify the second machine learning model; and
recognize the second geographical component from the captured destination information using the second machine learning model.

2. The system of claim 1 , wherein the plurality of machine learning models is configured to communicate data with an image database configured to store images of different first geographical components and different second geographical components for each of the different first geographical components.

3. The system of claim 2 , wherein the image database is further configured to store images of different third geographical components for each of the different second geographical components.

4. The system of claim 1 , wherein the plurality of machine learning models further comprises a third machine learning model trained to recognize a third geographical component from the captured destination information, wherein the one or more processors are further configured to:
based on recognizing the second geographical component, identify the third machine learning model; and
recognize the third geographical component from the captured destination information using the third machine learning model.

5. The system of claim 1 , wherein the one or more processors are configured to automatically instruct item processing equipment to process the item for delivery based on at least the first geographical component and the second geographical component.

6. A method of sorting items, the method comprising:
capturing, at an optical scanner, destination information provided on an item, the destination information comprising a first geographical component and a second geographical component;
storing, in a memory, a plurality of machine learning models, wherein the plurality of machine learning models comprises:
a first machine learning model trained to recognize the first geographical component from the captured destination information, and
a second machine learning model trained to recognize the second geographical component from the captured destination information based at least in part on the recognized first geographical component;

recognizing, by one or more processors, the first geographical component from the captured destination information using the first machine learning model;
based on the recognizing the first geographical component, identifying, by the one or more processors, the second machine learning model; and
recognizing, by the one or more processors, the second geographical component from the captured destination information using the second machine learning model.

7. The method of claim 6 , further comprising processing the item in item processing equipment for delivery to be based on at least the recognized first geographical component and the recognized second geographical component.

8. The method of claim 6 , wherein the plurality of machine learning models is in communication with an image database configured to store images of different first geographical components and different second geographical components for each of the different first geographical components.

9. The method of claim 8 , wherein the image database is further configured to store images of different third geographical components for each of the different second geographical components.

10. The method of claim 6 , further comprising:
based on the recognizing the second geographical component, identifying, by the one or more processors, a third machine learning model; and
recognizing, by the one or more processors, a third geographical component from the captured destination information using the third machine learning model.

11. A method of automatically recognizing geographical area information provided on an item, the method comprising:
capturing, at an optical scanner, geographical area information on an item, the geographical area information comprising a first geographical component and a second geographical component;
running a first one or more machine learning models, the first one or more machine learning models trained to recognize the first geographical component;
subsequent to running the first one or more machine learning models, identifying a second one or more machine learning models, the second one or more machine learning models trained to recognize the second geographical component; and
running the second one or more machine learning models.

12. The method of claim 11 , wherein the geographical area information further comprises a third geographical component.

13. The method of claim 12 further comprising, subsequent to running the second one or more machine learning models:
identifying a third one or more machine learning models, the third one or more machine learning models trained to recognize the third geographical component; and
running the third one or more machine learning models.

14. The method of claim 11 , further comprising automatically processing the item for delivery based on at least the recognized first geographical component and the recognized second geographical component.

Assignees

• United States Postal Service

Inventors

• Simpson Ryan J

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06V30/416

  Extracting the logical structure, e.g. chapters, sections or page numbers; Identifying elements of the document, e.g. authors · CPC title

• G06V30/19173

  Classification techniques · CPC title

• G06F18/28

  Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries · CPC title